Pith. sign in

REVIEW 3 major objections 6 minor 38 references

UN-DETR: Promoting Objectness Learning via Joint Supervision for Unknown Object Detection

T0 review · 3 major / 6 minor · reviewed 2026-08-11 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗ Review JSON

Pith's one-line read Training an objectness score on both box overlap and classification confidence lets a transformer detector outperform prior methods at finding objects of unseen categories.

desk verdict Solid system paper with a real but incremental contribution; the SOTA margin is confounded with unsupervised pretraining and the attribution to the new IPS head doesn't survive the authors' own ablation. read the letter →

arxiv 2412.10176 v1 pith:VYTQNSOE submitted 2024-12-13 cs.CV

Haomiao Liu , Hao Xu , Chuhuai Yue , Bo Ma This is my paper · ORCID
classification cs.CV
keywords unknownobjectdetectionobjectnesslearninginstancepresencescorejointsupervisiontransformerdetectoropen-worldunsupervisedpretrainingone-to-manyassignment
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

The paper argues that prior unknown-object detectors learn objectness from only one source of supervision, either box overlap or classification confidence, and that this split explains their poor recall or precision on unseen objects. It proposes UN-DETR, a transformer-based detector that trains a dedicated Instance Presence Score (IPS) under joint supervision from both spaces: a generalized-IoU term from the regression head and the summed foreground logits from the classification head. This score is then reused in query selection, in an IPS-guided NMS, and in a dual-criteria rule that separates known from unknown objects. On the COCO-OOD benchmark the method raises unknown F1 and precision by 7.0 and 11.2 points over the previous best, while roughly preserving known-class mAP, which is what the authors offer as evidence that the joint-supervision strategy is the cause.

What carries the argument

Instance Presence Score (IPS), a learnable scalar output by a dedicated one-layer IPS Predictor head, is the load-bearing object. Its training target is $P_o(e_i)=\alpha\,\mathrm{GIoU}(b_i,\hat{b}_{\sigma_{pos}})+\beta\,\hat{P}^f_{\sigma_{pos}}$, with $\alpha=0.6$, $\beta=0.4$; the first term is the generalized IoU between the regression head's box and the matched ground-truth box, and the second is the sum of the known-class logits, a foreground probability. Samples whose $\mathrm{GIoU}$ falls at or below $\tau=0.6$ are pushed toward a constant $C=0.5$ instead, giving the score a clean separation between foreground-like and background-like boxes. IPS is reused in three places: a one-to-many assignment adds a second set of matched queries as extra positives for IPS training only, Unbiased Query Selection replaces the two-stage DETR query-scoring classification head with an additional IPP head, and an IPS-guided post-process ranks boxes with Distance-IoU NMS before applying a dual-criteria rule (classification confidence and IPS both above thresholds $\Rightarrow$ known; low classification but high IPS $\Rightarrow$ unknown).

What would settle it

The claim would be falsified if re-running the released model on COCO-OOD while sweeping the dual-criteria thresholds reveals that the reported 54.9 U-F1 and 54.5 U-PRE are only obtained at the sweep maximum rather than at a pre-fixed threshold setting.

Watch

Extended reading notes

Core claim

The central discovery is that objectness for unknown objects is better learned as a recoupling of two complementary latent spaces rather than from either alone. The paper shows that using only categorical information yields clear foreground/background separation but poor instance separation, while using only positional information yields the opposite. IPS is defined as the probability of an object's presence and is trained to match a convex combination of GIoU(box, matched GT) and the sum of classification logits over known categories, with a low constant target for low-overlap samples. The same IPS is then used to select decoder queries and to guide post-processing, including a known/unknown decision rule based on both classification confidence and IPS. Under the UOD evaluation protocol, this design claims the highest reported unknown-object F1 and precision on COCO-OOD and COCO-Mixed while keeping known-object mAP roughly unchanged.

Load-bearing premise

The central premise is that an objectness score trained as a weighted blend of box overlap (GIoU) and the summed known-class logits transfers to unseen categories, and that the known/unknown thresholds in the final protocol are fixed rather than tuned to each test set.

Editorial extensions

If this is right

  • If the reported gains hold, augmenting any detector that has separate classification and regression heads with a jointly supervised IPS head should improve unknown-object recall and precision without retraining the backbone from scratch.
  • Because IPS is class-agnostic, using it in query selection should make the decoder's initial queries less biased toward known categories, which is a direct design lesson for open-world detectors.
  • The ablation showing U-PRE drops from 54.5 to 35.0 without unsupervised pretraining implies that an objectness prior acquired from region proposals is a major share of the method's precision advantage.
  • The dual-criteria rule (known if both classification confidence and IPS are high, unknown if only IPS is high) provides a template for separating known from unknown instances using two complementary confidence signals.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

  • A natural stress test the paper does not run is a full precision-recall sweep of the dual-criteria thresholds; if the reported COCO-OOD numbers sit at the peak of that sweep, the benchmark gains are partly tuning, not pure generalization.
  • The same joint-supervision recipe could be dropped into open-vocabulary detectors as an auxiliary objectness head, so that objects outside the prompt vocabulary still get localized and marked unknown rather than missed.
  • The COCO-Mixed results are more modest than the COCO-OOD ones: the method leads in U-AP and U-PRE but trails on U-F1, so a careful reader should weigh how much of the 'surpassing previous methods' claim is specific to the all-unknown test set.
  • Replacing the fixed convex weights $\alpha=0.6$, $\beta=0.4$ with a learned gating of the two signals is a testable variant that might reduce sensitivity to the choice of known categories.
Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Referee Report

3 major / 6 minor

Summary. The paper proposes UN-DETR, a transformer-based unknown object detector built on Deformable DETR. Its central novelty is an Instance Presence Score (IPS) predicted by a dedicated head, trained with joint supervision from positional (GIoU) and categorical (sum of known-class logits) signals, with a one-to-many assignment for additional positive samples. IPS is also used for query selection and post-processing, and the whole detector is unsupervised-pretrained on the training set with region priors. The paper reports state-of-the-art results on COCO-OOD and strong results on COCO-Mixed, and includes ablations of each component.

Significance. If the reported results hold under a fair comparison, this is a meaningful contribution: it is the first transformer-based UOD method, it introduces a learnable objectness score that combines two complementary signals, and it provides a fairly complete ablation study with convergence and visualization analyses. The authors also provide a code repository, which supports reproducibility. However, the central SOTA claim is currently supported by a comparison in which the proposed method uses an unsupervised pretraining recipe that is not applied to the baselines in the main tables, and the paper's own ablation shows the pretraining is largely responsible for the headline U-F1 and U-PRE margins. The contribution is therefore promising but not yet convincingly established.

major comments (3)
  1. [Ablation Study, Table 3 row 8] The claim that UN-DETR surpasses the previous leading method (Table 1: U-F1 54.9 vs 47.9, U-PRE 54.5 vs 43.3 on COCO-OOD) is confounded by the unsupervised pretraining component. Removing pretraining drops U-F1 from 54.9 to 44.5 and U-PRE from 54.5 to 35.0, both below UnSniffer's 47.9 and 43.3. The appendix's retrained-baseline results are not integrated into Tables 1-2, and the reported UnSniffer-with-pretraining U-AP of 43.9 is below its original 45.4. Thus the headline margin is not attributable to the jointly supervised IPS mechanism alone; please integrate the pretrained-baseline comparisons into the main tables, or provide a no-pretraining comparison against all baselines on the full metric set.
  2. [IPS-Guided Post Process, 'Dual-Criteria Unknown Distinguish Protocol'] The known/unknown decision is made by checking whether 'both classification confidence and IPS are above set thresholds', but the threshold values are never reported, and no sensitivity analysis or protocol for choosing them is given. Since U-F1 and U-PRE are computed after this decision rule, the headline results are not independently reproducible from the text. Please report the thresholds, state how they were selected (e.g., fixed in advance or tuned on COCO-OOD), and provide a sensitivity study over these thresholds.
  3. [Instance Presence Score Predictor, Eqs. (1)-(4)] The IPS supervision target Po = α·eo_bbox + β·eo_cls is not an independent objectness measurement: eo_cls is the sum of the model's own classification logits for the matched prediction, and eo_bbox is the GIoU between the model's predicted box and the ground-truth box. Using the model's own confidence as a categorical supervision signal is a legitimate training strategy, but the paper should clarify this explicitly and justify why it does not induce confirmation bias in the objectness estimate. A concrete validation would be to compare against an IPS trained with binary foreground/background labels for known objects, or to report the correlation between the learned IPS and GIoU on held-out known objects.
minor comments (6)
  1. [Table 1 caption] The caption writes 'COCO-ODD'; this should be 'COCO-OOD'.
  2. [Introduction, contributions] The text says 'jointly supervised singals from complementary positional and categorical latent space'; 'singals' should be 'signals', and 'latent space' should be plural.
  3. [Appendix, Unsupervised Pre-training] 'Futhermore' should be 'Furthermore'.
  4. [Eqs. (1)-(4)] The metric is written both as 'GIoU' and 'GIOU'; please unify the notation.
  5. [Unbiased Query Selection] The text says 'we introduce an additional IPP' for query selection, but the same acronym IPP is also used for the main objectness head. Please clarify whether these are the same head, separate heads, or shared-parameter heads, and how the two IPP losses interact.
  6. [Appendix, Qualitative Analysis] The appendix refers to 'the donut in the second row ... of Figure 3' and 'the moon in the fifth row of Figure 4'; these should reference the qualitative comparison figures (Figures 8 and 9).

Circularity Check

1 steps flagged · score 4.0 of 10

IPS is trained on the model's own summed classification logits and then used as an independent signal to reclassify low-confidence boxes; the core SOTA claim remains empirical and externally benchmarked.

  1. self definitional [Section 'Instance Presence Score Predictor' (Eqs. 1-5) and Section 'IPS-Guided Post Process' ('Dual-Criteria Unknown Distinguish Protocol')]
    "ˆeo_cls,σpos = ˆPf σpos (2) where Pf is the sum of K logits ... we set the objective probability Po(ei) = α · eo_bbox + β · eo_cls, which serves as the supervised signal for IPP training ... If both classification confidence and IPS are above set thresholds, the object is assigned to a known category. If classification confidence is low but IPS is high, the object is recognized but not confidently categorized, hence it's classified as unknown."

    The IPS regressor is trained to reproduce a target whose categorical component is literally the sum of the model's own classification logits. The dual-criteria unknown/known decision then treats 'IPS high' as though it were independent evidence that can override low classification confidence and label a box unknown. Since IPS is partly a learned copy of the classification head's summed logits, the decision 'classification confidence low but IPS high' reduces, by construction, to re-thresholding the model's own confidence signal rather than consulting a genuinely separate objectness measurement.

full rationale

The paper's headline result is an empirical state-of-the-art claim on public UOD benchmarks, not a mathematical derivation, so most of the argument is not circular in the formal sense. The one concrete reduction-by-construction is in the IPS design: Eq. (2) defines the categorical supervision as the sum of the classification head's own logits, and the post-processing protocol then uses IPS to make known/unknown decisions that are supposed to be independent of classification confidence. That is a partial self-definitional loop, but it does not collapse the whole method because the positional/GIoU component and the unsupervised pretraining are external ingredients. The larger validity concern is experimental: Table 3 row 8 shows that removing unsupervised pretraining drops U-F1 from 54.9 to 44.5 and U-PRE from 54.5 to 35.0, placing UN-DETR below UnSniffer on those metrics, while the appendix reports that retrained baselines with similar pretraining do not match the main-table numbers. This is a serious attribution and reproducibility problem, but it is a confound rather than a circularity. No load-bearing self-citation chain or imported uniqueness theorem appears in the paper. Overall score 4 reflects one partial self-definitional step in the IPS/post-process chain, with the central empirical comparison retaining independent content.

Assumptions & free parameters 6 free parameters · 5 assumptions · 2 invented entities

The central claim rests on the UOD task definition, on hand-set weights for the IPS target and loss, and on an unstated post-processing threshold. Several assumptions about GIoU and summed logits being valid objectness signals are domain assumptions with heuristic justification only.

free parameters (6)
  • alpha (positional weight in IPS target) = 0.6
    Weights GIoU-based positional supervision in the IPS target; set empirically with no sensitivity analysis.
  • beta (categorical weight in IPS target) = 0.4
    Weights the foreground-logit categorical supervision; complement of alpha, chosen empirically.
  • C (negative IPS constant) = 0.5
    Constant target for low-GIoU queries in Eq. 4; claimed to improve stability but no ablation is shown.
  • tau (GIoU positive/negative threshold) = 0.6
    Threshold in Eq. 3 and Eq. 4 dividing positive from negative IPP training samples; no sensitivity study.
  • loss weights lambda1, lambda2, lambda3 = 3, 2, 5
    Weights in Eq. 6 for IPS, classification, and regression losses; chosen without reported tuning analysis.
  • dual-criteria thresholds for known/unknown decision = unstated
    IPS-guided post-process requires 'set thresholds' for classification confidence and IPS; values are never given, so the reported U-PRE and U-F1 depend on unspecified calibration.
assumptions (5)
  • standard math Deformable DETR bipartite matching and transformer decoder update provide a valid detection backbone.
    Used as the base architecture throughout the paper.
  • domain assumption GIoU between a predicted box and its matched ground truth is a class-agnostic objectness signal that transfers to unseen categories.
    Used to define positional latent supervision in Eq. 1.
  • domain assumption The sum of known-class logits represents the probability of foreground, independent of specific categories.
    Used to define categorical supervision in Eq. 2.
  • domain assumption Suboptimal one-to-many matched queries are still reliable positives for IPS learning.
    Motivates Eq. 7 and Eq. 8 and the one-to-many assignment strategy.
  • domain assumption Unsupervised pretraining with Selective Search boxes and SwAV embeddings transfers objectness priors to the full detector.
    Invoked in the section 'Unsupervised Pretraining with Objectness Priors'.
invented entities (2)
  • Instance Presence Score (IPS)
    purpose: Scalar objectness output of the IPP, used for query selection, NMS ranking, and known/unknown separation.
    It is an internal network output; its only support is the paper's own experiments, and no external falsifiable handle is specified.
  • Unbiased Query Selection (UQS) additional IPP head
    purpose: Replace the classification head for selecting top-K decoder queries.
    Another internal component; effectiveness is shown only through the paper's ablations and a scatterplot.

how reviews work

0 comments
Cite this review

Pith. "Pith review of UN-DETR: Promoting Objectness Learning via Joint Supervision for Unknown Object Detection." pith.science (2026). https://pith.science/paper/VYTQNSOE

@misc{pith2026241210176,
  author       = {Pith},
  title        = {Pith review of: UN-DETR: Promoting Objectness Learning via Joint Supervision for Unknown Object Detection},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/VYTQNSOE}},
  note         = {Machine review of arXiv:2412.10176}
}
read the original abstract

Unknown Object Detection (UOD) aims to identify objects of unseen categories, differing from the traditional detection paradigm limited by the closed-world assumption. A key component of UOD is learning a generalized representation, i.e. objectness for both known and unknown categories to distinguish and localize objects from the background in a class-agnostic manner. However, previous methods obtain supervision signals for learning objectness in isolation from either localization or classification information, leading to poor performance for UOD. To address this issue, we propose a transformer-based UOD framework, UN-DETR. Based on this, we craft Instance Presence Score (IPS) to represent the probability of an object's presence. For the purpose of information complementarity, IPS employs a strategy of joint supervised learning, integrating attributes representing general objectness from the positional and the categorical latent space as supervision signals. To enhance IPS learning, we introduce a one-to-many assignment strategy to incorporate more supervision. Then, we propose Unbiased Query Selection to provide premium initial query vectors for the decoder. Additionally, we propose an IPS-guided post-process strategy to filter redundant boxes and correct classification predictions for known and unknown objects. Finally, we pretrain the entire UN-DETR in an unsupervised manner, in order to obtain objectness prior. Our UN-DETR is comprehensively evaluated on multiple UOD and known detection benchmarks, demonstrating its effectiveness and achieving state-of-the-art performance.

Figures

Figures reproduced from arXiv: 2412.10176 by the authors.

Figure 1
Figure 1. Joint supervision for objectness learning [PITH_FULL_IMAGE:figures/full_fig_p001_1.png] view at source ↗
Figure 2
Figure 2. The overall architecture of UN-DETR Preliminary Two-Stage Deformable DETR Pipeline In D-DETR Pipeline, an input image I is processed by a backbone network to extract features, which are fed into an encoder using attention mechanisms to produce an enhanced feature sequence. In the decoder, Nquery object queries are updated through self-attention and cross-attention with the encoder’s output, leading to refined querie… view at source ↗
Figure 3
Figure 3. Visualizations of discriminability scores [PITH_FULL_IMAGE:figures/full_fig_p004_3.png] view at source ↗
Figures from the paper (8 more)
Figure 4
Figure 4. Figure 4: Visualization of classification scores and IPS for [PITH_FULL_IMAGE:figures/full_fig_p005_4.png]
Figure 5
Figure 5. Figure 5: Example results on COCO-OOD (first two rows) [PITH_FULL_IMAGE:figures/full_fig_p007_5.png]
Figure 6
Figure 6. Figure 6: Convergence curves for different assignment [PITH_FULL_IMAGE:figures/full_fig_p010_6.png]
Figure 7
Figure 7. Figure 7: t-SNE visualization of various classes’ hidden vec [PITH_FULL_IMAGE:figures/full_fig_p011_7.png]
Figure 8
Figure 8. Figure 8: Example results on COCO-OOD datasets. Detections are overlaid on known (yellow) and unknown (blue) objects. [PITH_FULL_IMAGE:figures/full_fig_p012_8.png]
Figure 9
Figure 9. Figure 9: Example results on COCO-Mixed datasets. Detections are overlaid on known (yellow) and unknown (blue) objects. [PITH_FULL_IMAGE:figures/full_fig_p013_9.png]
Figure 10
Figure 10. Figure 10: Example results on COCO-OOD datasets. Detections are overlaid on known (yellow) and unknown (blue) objects. [PITH_FULL_IMAGE:figures/full_fig_p014_10.png]
Figure 11
Figure 11. Figure 11: Example results with different number of class prompts on COCO-OOD datasets. Detections are overlaid on known [PITH_FULL_IMAGE:figures/full_fig_p014_11.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

38 extracted references · 18 canonical work pages

  1. [1]

    J.; Herzig, R.; Chechik, G.; Rohrbach, A.; Darrell, T.; and Globerson, A

    Bar, A.; Wang, X.; Kantorov, V.; Reed, C. J.; Herzig, R.; Chechik, G.; Rohrbach, A.; Darrell, T.; and Globerson, A. 2022. Detreg: Unsupervised pretraining with region priors for object detection. In Proceedings of the IEEE/CVF conference on computer vision and pattern recognition, 14605--14615

    work page 2022

  2. [2]

    Bendale, A.; and Boult, T. E. 2016. Towards Open Set Deep Networks. In 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR), 1563--1572

    work page 2016

  3. [3]

    Carion, N.; Massa, F.; Synnaeve, G.; Usunier, N.; Kirillov, A.; and Zagoruyko, S. 2020. End-to-end object detection with transformers. In European conference on computer vision, 213--229. Springer

    2020

  4. [4]

    Caron, M.; Misra, I.; Mairal, J.; Goyal, P.; Bojanowski, P.; and Joulin, A. 2020. Unsupervised learning of visual features by contrasting cluster assignments. Advances in neural information processing systems, 33: 9912--9924

    2020

  5. [5]

    Chen, Q.; Chen, X.; Wang, J.; Zhang, S.; Yao, K.; Feng, H.; Han, J.; Ding, E.; Zeng, G.; and Wang, J. 2023. Group detr: Fast detr training with group-wise one-to-many assignment. In Proceedings of the IEEE/CVF International Conference on Computer Vision, 6633--6642

    work page 2023

  6. [6]

    Denouden, T.; Salay, R.; Czarnecki, K.; Abdelzad, V.; Phan, B.; and Vernekar, S. 2018. Improving reconstruction autoencoder out-of-distribution detection with mahalanobis distance. arXiv preprint arXiv:1812.02765

    arXiv 2018

  7. [7]

    Du, X.; Wang, X.; Gozum, G.; and Li, Y. 2022 a . Unknown-Aware Object Detection: Learning What You Don't Know from Videos in the Wild. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 13678--13688

    work page 2022

  8. [8]

    Du, X.; Wang, Z.; Cai, M.; and Li, Y. 2022 b . Vos: Learning what you don't know by virtual outlier synthesis. arXiv preprint arXiv:2202.01197

    arXiv 2022

Show all 38 references
  1. [9]

    K.; Winn, J.; and Zisserman, A

    Everingham, M.; Van Gool, L.; Williams, C. K.; Winn, J.; and Zisserman, A. 2010. The pascal visual object classes (voc) challenge. International journal of computer vision, 88: 303--338

    2010

  2. [10]

    S.; and Shah, M

    Gupta, A.; Narayan, S.; Joseph, K.; Khan, S.; Khan, F. S.; and Shah, M. 2022. Ow-detr: Open-world detection transformer. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 9235--9244

    2022

  3. [11]

    He, K.; Zhang, X.; Ren, S.; and Sun, J. 2016. Deep residual learning for image recognition. In Proceedings of the IEEE conference on computer vision and pattern recognition, 770--778

    2016

  4. [12]

    Hendrycks, D.; and Gimpel, K. 2017. A Baseline for Detecting Misclassified and Out-of-Distribution Examples in Neural Networks. In International Conference on Learning Representations

    2017

  5. [13]

    Jiang, B.; Luo, R.; Mao, J.; Xiao, T.; and Jiang, Y. 2018. Acquisition of localization confidence for accurate object detection. In Proceedings of the European conference on computer vision (ECCV), 784--799

    2018

  6. [14]

    S.; and Balasubramanian, V

    Joseph, K.; Khan, S.; Khan, F. S.; and Balasubramanian, V. N. 2021. Towards open world object detection. In Proceedings of the IEEE/CVF conference on computer vision and pattern recognition, 5830--5840

    2021

  7. [15]

    Krizhevsky, A.; Sutskever, I.; and Hinton, G. E. 2012. Imagenet classification with deep convolutional neural networks. Advances in neural information processing systems, 25

    2012

  8. [16]

    M.; and Zhang, L

    Li, F.; Zhang, H.; Liu, S.; Guo, J.; Ni, L. M.; and Zhang, L. 2022 a . Dn-detr: Accelerate detr training by introducing query denoising. In Proceedings of the IEEE/CVF conference on computer vision and pattern recognition, 13619--13627

    2022

  9. [17]

    H.; Zhang, P.; Zhang, H.; Yang, J.; Li, C.; Zhong, Y.; Wang, L.; Yuan, L.; Zhang, L.; Hwang, J.-N.; Chang, K.-W.; and Gao, J

    Li, L. H.; Zhang, P.; Zhang, H.; Yang, J.; Li, C.; Zhong, Y.; Wang, L.; Yuan, L.; Zhang, L.; Hwang, J.-N.; Chang, K.-W.; and Gao, J. 2022 b . Grounded Language-Image Pre-Training. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), 10965--10975

    2022

  10. [18]

    Liang, S.; Li, Y.; and Srikant, R. 2018. Enhancing The Reliability of Out-of-distribution Image Detection in Neural Networks. In International Conference on Learning Representations

    2018

  11. [19]

    Liang, W.; Xue, F.; Liu, Y.; Zhong, G.; and Ming, A. 2023. Unknown Sniffer for Object Detection: Don't Turn a Blind Eye to Unknown Objects. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 3230--3239

    2023

  12. [20]

    Liu, S.; Zeng, Z.; Ren, T.; Li, F.; Zhang, H.; Yang, J.; Jiang, Q.; Li, C.; Yang, J.; Su, H.; Zhu, J.; and Zhang, L. 2023. Grounding DINO: Marrying DINO with Grounded Pre-Training for Open-Set Object Detection. arXiv:arXiv:2303.05499

    2023 arXiv

  13. [21]

    Liu, W.; Wang, X.; Owens, J.; and Li, Y. 2020. Energy-based out-of-distribution detection. Advances in neural information processing systems, 33: 21464--21475

    2020

  14. [22]

    H.; Liu, H.; and Lv, F

    Ma, S.; Wang, Y.; Wei, Y.; Fan, J.; Li, T. H.; Liu, H.; and Lv, F. 2023. Cat: Localization and identification cascade detection transformer for open-world object detection. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 19681--19690

    2023

  15. [23]

    Neubeck, A.; and Van Gool, L. 2006. Efficient non-maximum suppression. In 18th international conference on pattern recognition (ICPR'06), volume 3, 850--855. IEEE

    2006

  16. [24]

    Redmon, J.; Divvala, S.; Girshick, R.; and Farhadi, A. 2016. You only look once: Unified, real-time object detection. In Proceedings of the IEEE conference on computer vision and pattern recognition, 779--788

    2016

  17. [25]

    Ren, S.; He, K.; Girshick, R.; and Sun, J. 2015. Faster r-cnn: Towards real-time object detection with region proposal networks. Advances in neural information processing systems, 28

    2015

  18. [26]

    R.; Van De Sande, K

    Uijlings, J. R.; Van De Sande, K. E.; Gevers, T.; and Smeulders, A. W. 2013. Selective search for object recognition. International journal of computer vision, 104: 154--171

    2013

  19. [27]

    Vaswani, A. 2017. Attention is All You Need. arXiv preprint arXiv:1706.03762

    2017 arXiv

  20. [28]

    Wu, Y.; Zhao, X.; Ma, Y.; Wang, D.; and Liu, X. 2022 a . Two-branch objectness-centric open world detection. In Proceedings of the 3rd International Workshop on Human-Centric Multimedia Analysis, 35--40

    2022

  21. [29]

    Wu, Z.; Lu, Y.; Chen, X.; Wu, Z.; Kang, L.; and Yu, J. 2022 b . UC-OWOD: Unknown-classified open world object detection. In European Conference on Computer Vision, 193--210. Springer

    2022

  22. [30]

    Yang, S.; Sun, P.; Jiang, Y.; Xia, X.; Zhang, R.; Yuan, Z.; Wang, C.; Luo, P.; and Xu, M. 2022. Objects in Semantic Topology. In International Conference on Learning Representations

    2022

  23. [31]

    Zhao, X.; Ma, Y.; Wang, D.; Shen, Y.; Qiao, Y.; and Liu, X. 2023. Revisiting open world object detection. IEEE Transactions on Circuits and Systems for Video Technology

    2023

  24. [32]

    Zheng, J.; Li, W.; Hong, J.; Petersson, L.; and Barnes, N. 2022. Towards open-set object detection and discovery. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 3961--3970

    2022

  25. [33]

    Zheng, Z.; Wang, P.; Liu, W.; Li, J.; Ye, R.; and Ren, D. 2020. Distance-IoU loss: Faster and better learning for bounding box regression. In Proceedings of the AAAI conference on artificial intelligence, volume 34, 12993--13000

    2020

  26. [34]

    Zhu, X.; Su, W.; Lu, L.; Li, B.; Wang, X.; and Dai, J. 2020. Deformable detr: Deformable transformers for end-to-end object detection. arXiv preprint arXiv:2010.04159

    2020 arXiv

  27. [35]

    Zohar, O.; Wang, K.-C.; and Yeung, S. 2023. Prob: Probabilistic objectness for open world object detection. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 11444--11453

    2023

  28. [36]

    Zong, Z.; Song, G.; and Liu, Y. 2023. Detrs with collaborative hybrid assignments training. In Proceedings of the IEEE/CVF international conference on computer vision, 6748--6758

    2023

  29. [37]

    , " * write output.state after.block = add.period write newline

    ENTRY address archivePrefix author booktitle chapter edition editor eid eprint howpublished institution isbn journal key month note number organization pages publisher school series title type volume year label extra.label sort.label short.list INTEGERS output.state before.all...

  30. [38]

    write newline

    " write newline "" before.all 'output.state := FUNCTION n.dashify 't := "" t empty not t #1 #1 substring "-" = t #1 #2 substring "--" = not "--" * t #2 global.max substring 't := t #1 #1 substring "-" = "-" * t #2 global.max substring 't := while if t #1 #1 substring * t #2 gl...

Pith tools

Reviewed August 11, 2026 · model on record in the stance chip above.